Background: The delivery and outcomes of care can be improved by using innovations (i.e., new knowledge, technologies and practices) supported by scientific evidence. Yet, moving evidence into routine care is not simple. This means many patients do not receive the best care possible.

Objectives: To examine the role of scientific evidence when adopting complex innovations, including how scientific evidence is considered amongst other relevant factors.

Methods: Using case study methodology, we studied the adoption (i.e. decision-making processes) of complex innovations in cancer care. Five cases were purposefully sampled from one health region in Nova Scotia, Canada, representing innovations with differing evidentiary bases. Data were collected using key informant interviews (n=32) and document (n>100) analysis. Analysis involved in-depth analyses of each case and a cross-case analysis, with findings compared and contrasted across cases. Three researchers coded and analysed the data, with research team meetings to question and clarify findings.

Results: Cross-case analysis revealed: 1) scientific evidence played a greater role in early adoption processes when clinicians/divisions decided to advocate for the innovation – but a limited role in subsequent decision making, with organisational decision makers trusting those who brought the innovation forward. In all cases, 2) evidence from a plurality of sources informed decision-making, including clinical experience, local data, and information from other jurisdictions. The latter provided important insight into implementation challenges and real-world impact. A range of evidence sources was sought given that 3) decision makers negotiated three key issues when making decisions: expected budgetary and operational implications; expected impact on patients; and equitable access to care. Finally, 4) steadfast leadership, mostly by frontline staff without formal leadership roles, was essential to eventual adoption.

Conclusions: Scientific evidence has a declining role as adoption decisions progress to higher levels of an organisation; scientific evidence was neither sufficient nor necessary for adoption.